We derive CT2 and Cq2 from high-resolution (grid spacing of 2-4 m) large-eddy simulations (LES) of a homogeneously-heated convective boundary layer, driven by measurements at Cabauw (The Netherlands). Three different methods to obtain the structure parameters from LES are investigated. Among others, it is found that the shape of the vertical profiles of the structure parameters compares well with former experimental and modelling results. Depending on the method, deviations in the magnitude up to a factor of two are found and traced back to effects of subgrid-scale modeling in LES.
In the presentation we compare the structure parameter profiles with data from an LAS ranging over a distance of 9.8km as well as aircraft measurements along the LAS path in different heights during RECAB 2002 (courtesy of Fred Bosveld at KNMI and courtesy of Beniamino Gioli at Ibimet). In the mixed layer we find that Cq2 compares well between LES and aircraft data, whereas CT2, derived from aircraft data, underestimates the proposed decrease with height. In the surface layer we find a good agreement with the LAS data. Virtual scintillometer measurements are employed in the LES in order to investigate the statistical error in LAS measurements due to the temporal and spatial variability of the path-averaged CT2 as well as its variability along the path.
Furthermore we present results from investigations of the similarity relationships for structure parameters in the surface layer. Universal functions, relating the structure parameters to the surface fluxes of sensible and latent heat, by means of Monin-Obkhukov similarity and local free convection similarity are calculated from the LES data and compared with the proposed formulations in literature.